Fastened Joint Analysis and Test Correlation of the MLA Beam Expander

Craig L. Stevens Craig.Stevens@nasa.gov Mechanical Systems Analysis & Simulation Branch NASA Goddard Space Flight Center Greenbelt, Maryland FEMCI Workshop Presentation May 6, 2004

Overview
1. 2. 3. 4. 5. 6. Introduction Problem Description Pre-Test Analysis Test Results Model Correlation Conclusions

Introduction
• Mercury Laser Altimeter (MLA) on Mercury Surface, Space Environment, Geochemistry, and Ranging (MESSENGER) spacecraft • MLA produces accurate measurements of topography and measures Mercury's wobble (due to the planet's libration) • MESSENGER Objectives:
– Determine the structure of Mercury’s mantle and crust – Investigate Mercury's polar caps – Determine the state of Mercury’s core (fluid or solid?)

• MESSENGER developed by Johns Hopkins University Applied Physics Laboratory (APL) • MLA developed by NASA Goddard Space Flight Center (Delivered to APL June 2003) • Launch
– August 2004 – Delta II 2925H-9-5

Introduction
MESSENGER Configuration
Gamma-Ray Spectrometer

MLA
Plasma Spectrometer
Neutron Spectrometer Energetic Particles Spectrometer

Atmospheric and Surface Spectrometer

Dual Imager on Pivot Platform

X-Ray Spectrometer

Introduction
MLA Configuration
Z

Receiver tubes [4] (Be)
Laser beam expander (Be) Laser bench (Be)

Y X

Main housing (Be)
Mounting flexures [3] (Ti) Aft optics (Be)

Power Converter Assy.
Housing (Mg)

Problem Description
Flight laser bench with attached beam expander •Three point mount: #2-56 fasteners •Beam expander: mass ~0.4 lbm, height ~6.5”

MLA laser workmanship level vibration test/analysis •Low level: 0.04 G2/Hz, 6.8 GRMS
Task: •Pre Test Analysis: Predict natural frequencies and loads •Correlate FEM with test data •Investigate test anomalies and determine cause •FEM: 80002 80001 •CQUAD4 model •Fixed Base
V1 C1

80003
Z X Y 123456

Pre-Test Analysis
•Initial joint stiffness estimated using tension joint theory Bolt:
Kb  Eb AEff Lb
Load, #

Joint Diagram 700 600 500 400 300 200 100 0 0.0E+00

Kb= 1.2E+6 #/in
Bolt Fitting Preload

E f D tan(a ) Fitting: K f  ( L tan(a )  d w  D)(d w  D) 2 ln ( L tan(a )  d w  D)(d w  D)

Preload = 250#

where, Eb = Bolt Young’s modulus Aeff = Bolt effective area Lb = Grip Ef = Fitting Young’s modulus D = Bolt diameter L = Clamped fitting thickness dw = Washer diameter a = Apex angle (assume 30o) •Bolt carries 10% of applied load

Kf=1.1E+7#/in
1.0E-04 2.0E-04 3.0E-04 Deflection, in 4.0E-04 5.0E-04 6.0E-04

dw

0.1” thick flange
L Lb

0.5” thick bench
D

#2-56 Fastener

Pre-Test Analysis
•Use Springs (CELAS) to model joint stiffness: Fitting stiffness = Kf/4 (9) Bolt stiffness = Kb (3) Normal constraints (center flange) = KNC (3) Normal constraints (near bolt) = KNCb (6) •Normal constraint locations chosen based on aluminum engineering model test data •Value of KNC and KNCb chosen based on aluminum engineering model test data and fitting stiffness •Same order of magnitude as fitting stiffness •1E+7 #/in •X and Y shear forces reacted through bolts using RBE2

2
Y
X

1 3

Pre-Test Analysis
Normal modes analysis:
MODAL EFFECTIVE MASS FRACTION: PRETEST
T1 T2 T3 MODE FREQ NO. (Hz) FRACTION SUM FRACTION SUM FRACTION SUM 1 509.8 2.14E-01 2.14E-01 4.65E-05 4.65E-05 6.76E-05 6.76E-05 2 518.3 4.05E-05 2.14E-01 2.13E-01 2.13E-01 3.77E-09 6.76E-05

Test Results
Z-axis: Test performed with no issues (no major modes) Y-Axis: Modal Frequencies Sine Signature Y
Item Analysis PreSine Y X Fn (Hz) Y Fn (Hz) 510 507 518 522

102

Y-Axis Pre Sine FRF

102

Y-Axis Pre Sine FRF

101 101 H12(f) 100 H22(f) 10-1 10-1 10
-2

100

10-3

200 400 600 800 1000 1200 1400 1600 1800 2000 Frequency, Hz

10-2 200 400 600 800 1000 1200 1400 1600 1800 2000 Frequency, Hz

Test Results
Y-axis (cont’):

•Frequency shift as random vibration levels increased:
•Harmonics observed in random beam expander responses •Integer multiples of resonant frequency •Indicates nonlinear response •Not present in sine runs
ASD (G2/Hz)

Axis Y

Frequency Migration X Fn (Hz) Y Fn Test Pre Sine 507 -18 dB 500 -6 dB 444 Full Level 402 Post Sine 482

(Hz) 522 502 440 410 451

Y-Axis Laser Random Vibration Workmanship -18 dB Level
10 Beam Expander Y Response, 7.765 Grms Laser Bench Y Response, 1.055 Grms Controller Input, 0.853 Grms -18 dB Input Spec, 0.854 Grms

1

0.1

0.01

0.001

0.0001

0.00001

0.000001 10 100 Frequency (Hz) 1000 10000

Test Results
Why do harmonics of resonant frequency exist?: •Asymmetrical distortion of sine wave

•Several reasons: Contact Boundary Condition

Test Results
•Frequency shifts observed in pre and post sine sweeps •Modal frequencies drop over 10% Modal Frequencies X F (Hz) Item •X and Y modes switch order
n

Y Fn (Hz) 522 451

PreSine Y Post Sine

507 482

Y-Axis Sine FRF 10
2

Y-Axis Sine FRF
Pre Sine Y Post Sine Y

10

2

Pre Sine Y Post Sine Y

101 101 100

H12(f)

H22(f) 200 400 600

10-1 10-2

100

10-1 10-3 10-4
-2 800 1000 1200 1400 1600 1800 2000 10 200 Frequency, Hz

400

600

800 1000 1200 1400 1600 1800 2000 Frequency, Hz

Model Correlation
•Use MSC NASTRAN Solution 200 to correlate model and investigate frequency shift •Assume linear modes and mode shapes •Vary CELAS stiffnesses to investigate preload loss •Key cards: •MODTRAK: Performs cross orthagonality check to ensure proper modes meet constraints •DEQATN: Create Least squares objective function formulation

2
Y X

1 3

E ( f n1 , f n 2 , f n1act , f n 2 act , W1 , W2 )  W1 ( f n1  f n1act ) 2  W2 ( f n 2  f n 2 act ) 2
•For Post Sine Y correlation: W1 = 1.0 W2 = 1.0 fn2act = 451 Hz (Y Mode) fn1act = 482 Hz (X Mode)

Model Correlation
Results:
$ UPDATED DESIGN MODEL DATA ENTRIES $ DESVAR * 1NCS *D 1V 1.00000000E+10 DESVAR * 2KB1 *D 2V 1.00000000E+10 DESVAR * 3KF1 *D 3V 1.00000000E+10 DESVAR * 4KB2 *D 4V 1.00000000E+10 DESVAR * 5KF2 *D 5V 1.00000000E+10 DESVAR * 6KB3 *D 6V 1.00000000E+10 DESVAR * 7KF3 *D 7V 1.00000000E+10 DESVAR * 8NCB1 *D 8V 1.00000000E+10 DESVAR * 9NCB2 *D 9V 1.00000000E+10 DESVAR * 10NCB3 *D 10V 1.00000000E+10 7.72887313E+05 1.19993400E+06 9.05991600E+05 1.19993800E+06 3.37498700E+06 1.19594200E+06 1.00000000E+04+D 1.00000000E+04+D 1.00000000E+04+D 1.00000000E+04+D 1.00000000E+04+D 1.00000000E+04+D 1V 2V 3V 4V 5V 6V

1.45899362E+05
8.14678800E+06 4.54948100E+06 1.00000000E+04

1.00000000E+04+D
1.00000000E+04+D 1.00000000E+04+D 1.00000000E+04+D

7V
8V 9V 10V

$ MODE TRACKING HAS BEEN PERFORMED SUCCESSFULLY WITH SOME CHANGE IN THE MODE ORDER. $ ALL THE DRESP1 ENTRIES ASSOCIATED WITH MODE TRACKING ARE WRITTEN HERE FOR CONVENIENCE. $ DRESP1 101FN1 FREQ 2 DRESP1 102FN2 FREQ 1

Model Correlation
Results (Cont’):
Item NCS KB1 KF1 KB2 KF2 KB3 KF3 NCB1 NCB2 NCB3 Initial 1.00E+07 1.20E+06 3.00E+06 1.20E+06 3.00E+06 1.20E+06 3.00E+06 1.00E+07 1.00E+07 1.00E+07 Post % change 7.73E+05 -92.3% 1.20E+06 0.0% 9.06E+05 -69.8% 1.20E+06 0.0% 3.37E+06 12.5% 1.20E+06 0.0% 1.46E+05 -95.1% 8.15E+06 -18.5% 4.55E+06 -54.5% 1.00E+04 -99.9%

2
Y
X

1
3

Joint Diagram 700 600

Over 90 % stiffness loss in fastener 3 (or 2) •Preload ~0 •Results from post test investigation verified no structural failure and no other significant source for frequency shift

500 400 300 200 100 0 0.0E+00

Load, #

Bolt Fitting Preload

1.0E-04

2.0E-04

3.0E-04 Deflection, in

4.0E-04

5.0E-04

6.0E-04

Conclusions
• • • • Presence of harmonics in broadband responses indicative of nonlinearities in test data Optimizer is quick and efficient method to correlate modes and mode shapes of models Over 90% stiffness loss in fastener 2 or 3 Results indicate preload loss is the probable cause of frequency shift

References 1. Joseph Shigley. Mechanical Engineering Design. Fifth Edition. McGraw-Hill, Inc. New York. 1989 2. Thomas Sarafin. Design and Analysis of Fastened Joints for Aerospace Engineers. Course Notes. Instar Engineering and Consulting, Inc. 2001. 3. John H. Bickford. Handbook of Bolts and Bolted Joints. Marcel Dekker, Inc. New York 4. Ryan Simmons. MLA Laser Subassembly Workmanship Vibration Post-Test Report. NASA GSFC, Greenbelt, Maryland, 2003. MLA-RPT-000127 5. Steven W. Smith. The Scientist and Engineer’s Guide to Digital Signal Processing, 2nd Edition. California Technical Publishing, San Diego, California, 1999. 6. Scott Gordon. Private Correspondence. NASA GSFC, Greenbelt, Maryland. 7. Daniel Kaufman. Private Correspondence. NASA GSFC, Greenbelt, Maryland.

Sign up to vote on this title
UsefulNot useful